Sudden cardiac death (SCD) continues to pose a significant health care challenge with an annual incidence of over 350,000 and low survival rates. Implantable cardioverter defibrillators (ICDs) prevent SCD in patients with left ventricular (LV) systolic dysfunction. However, the critical survival benefit afforded by these devices is accompanied by short and long-term complications and a high economic burden. Moreover, in using current practice guidelines of LV ejection fraction (LVEF) d 35% as the main determining factor for patient selection, only a minority of patients actually benefit from ICD therapy (<25% in 5 years). There is an essential need for more robust diagnostic approaches to SCD risk stratification. This project examines the hypothesis that LV structural abnormalities are important independent predictors of SCD risk as they identify the presence and extent of the abnormal pathophysiologic substrate required for the ventricular arrhythmogenicity leading to SCD. This premise is supported by studies in pre-clinical models and limited patient cohorts examining the contribution of individual LV structural indices. However, there has been no prospective study of primary prevention ICD candidates in sufficiently large numbers to investigate the incremental value of a comprehensive, combined assessment of LV structure on SCD risk over and above that of LVEF and readily available demographic and clinical variables. We will utilize the largest, best characterized, current treatment era ICD patient population with a target study size of 400 patients and advanced cardiac magnetic resonance imaging to address this hypothesis. LV structure can be quantified in detail using cardiac magnetic resonance imaging with late gadolinium enhancement (CMR-LGE). Specifically, accurate assessment of LV volumes, mass, geometry, and infarct/scar characteristics, as well as global LV function, are feasible and obtainable clinically in a single examination. We aim to examine whether or not a clinical risk score incorporating individual CMR indices or a combination of indices are better able to discriminate between patients with high versus low susceptibility to SCD within the broader population of reduced LVEF patients. If the results of these studies demonstrate that metrics of LV structure are important independent prognostic risk factors, it may then be possible to more specifically target ICD therapy to those who are most likely to benefit and avoid unnecessary device implantations and their associated morbidities and costs in those who do not have an arrhythmogenic LV structural phenotype.